Software bug report dataset from Eclipse projects.

Area of Science:

• Software Engineering
• Empirical Software Engineering
• Data Science in Software Development

Background:

• Analysis of software project data (e.g., source control, defect tracking) enhances understanding of software development.
• Existing data extraction methods are time-consuming and data maintenance is often poor.
• Transitioning from Bugzilla to newer systems risks losing historical bug data.

Purpose of the Study:

• To introduce an extensive, up-to-date dataset of bug reports from the Eclipse bug-tracking system (Bugzilla).
• To provide a custom Command Line Interface (CLI) tool for efficient Bugzilla data extraction.
• To support research in defect prediction, software maintenance, and evolutionary analysis.

Main Methods:

• Generated a dataset comprising 301,378 bug reports and associated information from Bugzilla repositories for Eclipse products.
• Organized the dataset into project-specific folders for clarity.
• Developed a lightweight CLI tool to automate detailed data extraction from Bugzilla instances.

Main Results:

• Created the largest, most complete, and up-to-date publicly available dataset of Eclipse bug reports.
• The CLI tool enables efficient and automated data collection from diverse Bugzilla repositories.
• The dataset preserves historical bug information potentially lost during system migrations.

Conclusions:

• The presented dataset and CLI tool offer valuable resources for the software engineering research community.
• Facilitates advanced analyses in defect prediction, software maintenance, and understanding software evolution.
• Addresses the need for accessible and well-maintained historical software project data.